Type | |
Stats | 368 1,845 |
Reviews | (57) |
Published | Sep 14, 2023 |
Base Model | |
Training | Steps: 1,968 |
Usage Tips | Clip Skip: 1 |
Trigger Words | VendBecky VendOutfitWeddingDress VendOutfitScientist VendOutfitBSAA |
Hash | AutoV2 38F1F1849A |
This model is trained on screengrabs from the RE: Vendetta movie.
Basic usage and notes:
Use at 1.0 strength
Primary trigger word: VendBecky girl
I suggest covered with blood as part of the negative prompt as Rebecca has bloodstains on her face and clothing in some parts of the training set.
I also suggest teeth as part of the negative prompt if you want a facial expression that doesn't show teeth. Most of the training set has images of Rebecca talking so the model leans towards always showing her teeth
It was difficult getting shots of the lower parts of her outfits, so the AI will basically do guesswork when it comes to the bottom part of her wedding dress, or the pants of her other outfits.
Model was trained with clip skip 1, but I've found it to also work fine with clip skip 2.
Example triggers for specific outfits from the movie:
VendBecky girl wearing VendOutfitWeddingDress
VendBecky girl wearing green VendOutfitScientist and lab coat and glasses
VendBecky girl wearing green VendOutfitScientist and lab coat
VendBecky girl wearing green VendOutfitScientist
VendBecky girl wearing VendOutfitBSAA
VendBecky girl wearing a blanket
Check out the prompts on the preview images for full prompt examples.